|
Message-ID: <630D32C95CDF44FFBF083460739553CC@ath64dual> Date: Wed, 28 Oct 2009 23:19:59 -0500 From: "JFoug" <jfoug@....net> To: <john-users@...ts.openwall.com> Subject: Re: My patches (Performance, options, mingw, vc, generic md5, phpass, etc) ----- Original Message ----- From: "Solar Designer" <solar@...nwall.com> > >> macosx-x86-sse2: [..] >> This change is for MacOS 10.6 (Snow Leopard) ONLY. > > Does the addition of -m32 break anything on older versions of OS X? > I would expect it to be a no-op there, so I am thinking of making the > change in the official JtR tree. Might it be better to make a new rule, and leave that one. Then you could have a macosx32-x86-sse2 or something like that, since that is what you are specifying as the build, with the -m32's At some time, I am going to get a 64 bit OS installed again (I gave up on trying to get xp64). It is also getting the compiler and other things, to get the intrensics to build, or in other words, to get the md5 stuff I have been working on, to properly build with the intrisic SSE2 code. It is just up to this point, the only thing I have to develop with, is 32 bit MinGW, thus I have used OpenSSL's MD5 for all 'x86' or 'other OS' code, and used the md5-mmx.S (with the stages_*_md5.S). That builds good for 32 bit, but I think it is lacking for 64 bit SSE processors (I have been told). I am hopeful, that once I have time to get my work where it builds properly in x86 or 32 bit SSE/MMX or 64 bit SSE intrinsics, then build problems should subside, since the 'proper' build for the system (the macosx-x86-64 in this case), can be used. I just wish I was not so tied up right now with the 'real' job, eating all of my time. Jim.
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.